Electrical tests on PCB insulation materials and investigation of influence of solder fillets geometry on partial discharge

Transcription

1 , Firenze, Italy Electrical tests on PCB insulation materials and investigation of influence of solder fillets geometry on partial discharge A. Bulletti, L. Capineri B. Dunn ESTEC Material and Process Division EMPS-3 in Genoa, Italy, 21 March 2012 Research activities at Ultrasound and Non Destructive Testing Laboratory Design and fabrication of innovative or customized transducers Design of advanced ultrasonic electronic systems for real time signal and image processing Developments of diagnostic methods for medical and non destructive testing ultrasound Subsurface soil investigations with georadar, holographic radar and acousto-seismic methods Methods and instruments for the characterization of material properties Design of electronic equipments for industrial processes control AISEM Firenze -Italy 1

2 Sensors and instruments for NDT and industrial applications Lamb wave inspection on metal and composite laminate (ESA/ITI 2004) (Sx) Pyroelectric matrix sensor 8x8 elements For IR laser beam monitoring (Dx) Front end electronics with microcontroller programming Array of airborne ultrasound transducers at 120 khz with real-time electronics (National Semiconductor Corporation, 2011) Scope Partial discharge testing of solder fillets on PCB in partial vacuum: experimental results Characterization of electrical properties of printed circuit boards base laminates AISEM Firenze -Italy 2

3 Partial Discharge Testing of Solder Fillets on PCB in Partial Vacuum: Experimental Results This work investigates the influence on the partial discharge (PD) onset due to the geometry of solder fillets of aerospace and spacecraft component assemblies operating in vacuum and at high voltages. An automatic measuring system has been developed and calibrated to detect PD signals in vacuum. Two types of solder fillet geometries have been realized for each electrode configuration: round joints and sharp pointed joints. Corona Inception Voltage (CIV) and Corona Extinction Voltage (CEV) were recorded at different values of pressure x distance. PD detector and processing system Vacuum chamber: Diameter 45cm, Height 50 cm High Voltage (0-20kV) programmable DC Power Supply PD passive detector PD calibration and signal conditioning instrument Turbo-molecular vacuum pump, controllable pressure range: 10-2 mbar - 10 mbar. Software for PD acquisition/processing in real-time (LabView) 12bit / 1.25 MHz AISEM Firenze -Italy 3

4 Examples of test objects SOLDER FILLETS Sharp PADS GEOMETRY 6 mm Round 5 mm RESIN CALIBRATION BLOCK Real-time signal processing and analysis Example of real-time data output for the evaluation of the partial discharge activity before the inception of the glow discharge (CIV) on pairs of round shaped soldered fillet terminals on PCB at distance 6 mm. Round shaped soldered fillet Testing voltage : 530 Volt CIV: 545 V Pressure : 2 mbar Mesurement time interval: 5 minutes Typical discharge pulse Amplitude histogram Temporal events histogram AISEM Firenze -Italy 4

5 Concluding remarks Circular pads Sharp pads The improved sensitivity of the present laboratory set-up has confirmed that the corona inception voltage (CIV) of the solder joint configurations having sharp solder fillets have a Paschen curve minimum value of 550 V. This is only 50 V lower than the 600 V minimum determined for round, smooth solder joints. Concluding remarks cont ed Two electrode geometries with parallel copper lines and facing sharp corner electrodes, with solder terminals at a distance greater than 5 mm, have not revealed substantial differences in terms of PD activity (amplitude and temporal histograms) between the two types of solder fillets. This confirms the theoretical assumptions which predict that it is the minimum distance between electrodes which influences the CIV value and hence the PD activity. Deliberately contaminated joints (flux residue left without cleaning) confirmed that the very active RA flux (i.e. Alpha ) gave a far worse CIV than the residues from pure rosin (type R) when tested under vacuum between two adjacent solder joints. References: Capineri, L.; Dainelli, G.; Materassi, M.; Dunn, B.D, Partial discharge testing of solder fillets on PCBs in a partial vacuum: new experimental results, Electronics Packaging Manufacturing, IEEE Transactions onvolume: 26, Issue: 4, pp. 265, 2003 AISEM Firenze -Italy 5

6 Characterization of electrical properties of printed circuit boards base laminates New dielectric materials have been introduced for printed circuit board applications, such as Thermount (85NT) and Polyimide (N7001-1) with the aim to match the requirements for high speed and high density of electronic devices that are planned for new spacecraft electronic boards. The scope of this study is to report quantitative characterization of the surface and volume resistivity for the different material samples under various testing conditions that include relative humidity, electron beam irradiation, and UV exposure. Set-up for surface and volume resistivity measurements performed in clean room GPIB cable Triaxial cable Triaxial cable Electrometer Keithley 617 Resistivity Test Fixture Keithley 8009 High Voltage Source Keithley 237 Personal Computer Schematic for volume resistivity measurements Schematic for surface resistivity measurements Guard electrode Sample Sample A Guard electrode A V V AISEM Firenze -Italy 6

7 Characterization of as received samples: Surface and Volume Resistivity (R S and R V ) EPOXY FR4 POLYIMIDE THERMOUNT R S = 1.02E+16 Ω R S = 1.77E+17 Ω R S = 6.66E+16 Ω σ = 4.72E+15 Ω σ = 5.05E+16 Ω σ = 3.46E+16 Ω EPOXY FR4 POLYIMIDE THERMOUNT R V = 2.57E+16 Ωcm R V = 3.34E+16 Ωcm R V = 1.29E+16 Ωcm σ = 6.09E+15 Ωcm σ = 2.96E+15 Ωcm σ = 2.03E+15 Ωcm Each set counts 30 samples (thickness 1.6mm ±10%) Experimental set-up for UV exposure (λ = 365nm) Top view 300mm 70mm Intensity of illumination 2mW/cm 2 for 10 minutes (1.2J/cm 2 ) samples 30º Inclination of 30 to guarantee a uniform intensity on each sample AISEM Firenze -Italy 7

8 Summary of data of UV exposure tests Epoxy FR4: R S decreases of 28% Þ Polyimide: R S decreases of 14% Þ Thermount: R S decreases of 62% The surface resistivity variations remain within the standard deviation. However Thermount shows largest variation. Summary of data of E-beam irradiation tests Electron beam irradiation 10-3 mbar (297keV, 10mA for 7 minutes) Epoxy FR4: R V decreases of 93% (~1 order of magnitude) Þ Polyimide: R V decreases of 99.5% (>2 order of magnitude) Þ Thermount: R V decreases of 96% (>1 order of magnitude) The volume resistivity of base laminate materials decreases up to 2 order of magnitude when measured after 10 minutes from the end of the irradiation and recovers almost completely after 24 hours and completely after 15 days. AISEM Firenze -Italy 8

9 Set-up for relative humidity conditioning procedure Epoxy Polyimide Thermount Testing chamber conditions: T = 22 C±2ºC RH = (90±1)% Exposure time: 20 days Summary of results for RH exposure tests 90% RH exposure (20 T = (22±2)ºC) Epoxy - R S decreases of 70% Þ Polyimide - R S decreases of 98% (>1 order of magnitude) Þ Thermount - R S decreases of 96.5% (>1 order of magnitude) Þ Epoxy - R V decreases of 82.5% Þ Polyimide - R V decreases of 77% Thermount - R V decreases of 87% AISEM Firenze -Italy 9

10 Observations on RH exposure tests With 90% relative humidity tests at room temperature for 20 days the surface resistivity present a more evident decrease except for Epoxy samples. The recover effect after one month has been partially verified for Polyimide and Thermount. Further tests with relative humidity with higher temperature (90% 70ºC for 96 hours) to accomplish to ASTM standards Observations on results (90% 70ºC for 96 hours) Comparing the 90% 22ºC for 20 days exposure test with 90% 70ºC for 96 hours exposure test, we can observe that: Relative humidity tests with high temperature accelerate the process of variation of surface and volume resistivity. AISEM Firenze -Italy 10

11 Final Remarks Thermount showed a greater sensitivity to environmental conditions respect to a standard Epoxy and this lead to a larger variation of its electrical parameters. For both Polyimide and Thermount the larger decrease observed from their initial values are within two orders of magnitude but always greater than the minimum values declared by the manufacturer (E-24/125 and IPC-TM-650). References: A. Bulletti, L. Capineri, M. Materassi, Barrie D. Dunn, Surface resistivity characterisation of new printed circuit board materials for use in spacecraft electronics, IEEE transaction on Electronics and Packaging Manufacturing, vol. 30, n. 2, April, 2007 A. Bulletti, L. Capineri, M. Materassi & B. Dunn, Characterisation of Electrical Properties of Printed Circuit Board Base Laminates, Proceedings of 10th International Symposium on Materials in a Space Environment, Colliure, France, June 2006., Firenze, Italy Electrical tests on PCB insulation materials and investigation of influence of solder fillets geometry on partial discharge A. Bulletti, L. Capineri B. Dunn ESTEC Material and Process Division EMPS-3 in Genoa, Italy, 21 March 2012 AISEM Firenze -Italy 11